— / live dashboard / 567,528 rows / 94 markets
More titles, fewer hours each.
Netflix’s viewing keeps growing: 95.19 billion hours in H1 2025, 97.66 billion in H1 2026. But the catalogue is growing faster, so the average title earns less than it did. This is the dashboard the project was built for, over the same model, answering the questions a streaming engagement team would actually ask: what drives the hours, which kind of content earns them, and how long a hit holds a market.
Netflix What We Watched reports + weekly Top 10 · Python · SQL · SQLite
Loading the dashboard…
02 / where the numbers come from
Two public sources that don’t share a grain.
Netflix’s half-yearly What We Watched report lists every title with the hours it earned, about 99% of all viewing on the platform, twice a year and with no geography. Its weekly Top 10 has the weeks and the countries, back to 4 Jul 2021, but only ten titles per category and, for countries, only a rank.
So the project models them as two fact tables at their own grains, 50,028 title-by-half rows and 517,500 title-by-week-by-market rows, sharing date, title and region dimensions. The half-yearly totals reproduce the figures Netflix published for the same reports before anything was analysed. This page is built from that model by a script that reruns the project’s SQL, so a new data drop changes the numbers here rather than leaving a stale sentence behind.
03 / what it can’t tell you
Where the data runs out.
There are no hours by country.
Netflix publishes rank by country and nothing else, so every regional panel here is a stated proxy: how long titles stay in a chart, and how often a market’s number one is the world’s. Neither is an hour, and neither is presented as one.
Weekly hours only exist for hits.
The weekly lines are the global Top 10’s hours. They show the shape of what charted, not of everything people watched that week.
Netflix's rollups are not titles.
From the second half of 2025 the report folds its long tail into “Other Shows” and “Other Movies”. In H1 2026 “Other Shows” is the second-largest row in the file. They count towards the totals, which is how Netflix reports them, and are left out of every per-title view.
There are no genres.
Content type means series or film, English or not. Netflix doesn’t publish genre, so answering which genres over-index means matching titles against a second source.